Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2025 Jul 1;16(1):5653.
doi: 10.1038/s41467-025-60806-1.

Glial reactivity correlates with synaptic dysfunction across aging and Alzheimer's disease

Francieli Rohden  1   2 , Pamela C L Ferreira  1 , Bruna Bellaver  1 , João Pedro Ferrari-Souza  1   2 , Cristiano S Aguzzoli  1   3   4 , Carolina Soares  1   2 , Sarah Abbas  1 , Hussein Zalzale  1 , Guilherme Povala  1 , Firoza Z Lussier  1 , Douglas T Leffa  1 , Guilherme Bauer-Negrini  1 , Nesrine Rahmouni  5   6 , Cécile Tissot  1   5   6 , Joseph Therriault  5   6 , Stijn Servaes  5   6 , Jenna Stevenson  5   6 , Andrea L Benedet  7   8 , Nicholas J Ashton  7   8   9   10 , Thomas K Karikari  1   7   8 , Dana L Tudorascu  1   11 , Henrik Zetterberg  7   8   12   13   14   15 , Kaj Blennow  7   8 , Eduardo R Zimmer  2   3   16   17 , Diogo Souza  2 , Pedro Rosa-Neto  5   6 , Tharick A Pascoal  18   19
Affiliations

Glial reactivity correlates with synaptic dysfunction across aging and Alzheimer's disease

Francieli Rohden et al. Nat Commun. .

Abstract

Previous studies suggest glial and neuronal changes may trigger synaptic dysfunction in Alzheimer's disease (AD), but the link between their markers and synaptic abnormalities in the living brain remains unclear. We investigated the association between glial reactivity and synaptic dysfunction biomarkers in cerebrospinal fluid (CSF) from 478 individuals in cognitively unimpaired (CU) and cognitively impaired (CI) individuals. We measured amyloid-β (Aβ), phosphorylated tau (pTau181), astrocyte reactivity (GFAP), microglial activation (sTREM2), and synaptic markers (GAP43, neurogranin). CSF GFAP levels were associated with presynaptic and postsynaptic dysfunction, independent of cognitive status or Aβ presence. CSF sTREM2 levels were related to presynaptic markers in cognitively unimpaired and impaired Aβ+ individuals, and to postsynaptic markers in cognitively impaired Aβ+ individuals. Notably, CSF pTau mediated the relationships between GFAP or sTREM2 and synaptic dysfunction. Our findings, validated in two independent cohorts (TRIAD and ADNI), reveal a distinct pattern of glial contribution to synaptic degeneration.

PubMed Disclaimer

Conflict of interest statement

Competing interests: Therriault has served as a paid consultant for the Neurotorium educational platform and Alzheon. Ashton received payment for lectures from Biogen, BioArctic, Eli-Lilly, and Quanterix. Karikari reported receiving grants from the National Institutes of Health (NIH) and personal fees from the University of Wisconsin–Madison and the University of Pennsylvania outside the submitted work. Additionally, he holds a patent for WO2020193500A1. Zetterberg received personal fees from multiple companies, including AbbVie, Acumen, Alector, Alzinova, ALZPath, Amylyx, Annexon, Apellis, Artery Therapeutics, AZTherapies, Cognito Therapeutics, CogRx, Denali, Eisai, LabCorp, Merry Life, Nervgen, Novo Nordisk, Optoceutics, Passage Bio, Pinteon Therapeutics, Prothena, Red Abbey Labs, reMYND, Roche, Samumed, Siemens Healthineers, Triplet Therapeutics, and Wave. He also received personal fees for sponsored lectures from Alzecure, Biogen, Cellectricon, Fujirebio, Lilly, Novo Nordisk, and Roche outside the submitted work. Furthermore, Zetterberg is a cofounder of Brain Biomarker Solutions, part of the GU Ventures Incubator Program, outside the submitted work. Blennow has served as a consultant and on advisory boards for Acumen, ALZPath, BioArctic, Biogen, Eisai, Eli Lilly and Co, Moleac Pte Ltd, Novartis, Ono Pharma, Prothena, Roche Diagnostics, and Siemens Healthineers. He has also served on data monitoring committees for Julius Clinical and Novartis, given lectures, produced educational materials, and participated in educational programs for AC Immune, Biogen, Celdara Medical, Eisai, and Roche Diagnostics. Additionally, he is a cofounder of Brain Biomarker Solutions, part of the GU Ventures Incubator Program, outside the submitted work. Zimmer served on the Scientific Advisory Board (SAB) of Novo Nordisk, serves on the SAB of Next Innovative Therapeutics (Nintx), and is a cofounder of MASIMA. Rosa-Neto served on the SAB of Novo Nordisk, Eisai, and Eli Lilly, and acted as a consultant for Eisai and Cerveau Radiopharmaceuticals. The remaining authors declare no competing interests.

Figures

Fig. 1
Fig. 1. CSF GFAP is associated with synaptic dysfunction biomarkers independently of amyloid-β (Aβ) status or cognitive impairment.
Scatter plots illustrate significant positive associations between CSF GFAP and CSF GAP43 (a–f; presynaptic marker) as well as CSF neurogranin (Ng; g–l; postsynaptic marker) in cognitively unimpaired Aβ− (a, d, g, j), cognitively unimpaired Aβ+ (b, e, h, k), and cognitively impaired Aβ+ (c, f, i, l) individuals from the TRIAD (a–c, g–i) and ADNI (d–f, j–l) cohorts. Linear regression models were performed as two-sided tests and adjusted for age and sex. Shaded areas around the regression lines represent the standard error of the mean (SEM). Data points are color-coded by group: light pink for CU Aβ−, pink for CU Aβ+, and blue for CI Aβ+. The exact p value (1 f) is p = 1 × 10−9, and 1 l) is p = 8 × 10−10.
Fig. 2
Fig. 2. CSF sTREM2 is associated with synaptic dysfunction biomarkers only in the presence of amyloid-β (Aβ) pathology.
The scatter plots revealed that the associations between CSF sTREM2 and presynaptic marker CSF GAP43 were significant in CU Aβ+ and CI Aβ+ in TRIAD and ADNI cohorts (CU Aβ− a, d; CU Aβ+ b, e; CI Aβ+: c, f). CSF sTREM2 was associated with postsynaptic marker CSF Ng only in CI Aβ+ individuals in both TRIAD and ADNI cohorts (CU Aβ− g, j; CU Aβ + h, k; CI Aβ+: I, l). Shaded areas around the regression lines represent the standard error of the mean (SEM). Linear regression models were performed as two-sided tests and adjusted for age and sex. Data points are color-coded by group: light pink for CU Aβ−, pink for CU Aβ+, and blue for CI Aβ+. The exact p value (2e) is p = 2 × 10−5, and 2f) is p = 1 × 10−9.
Fig. 3
Fig. 3. CSF synaptic biomarker levels increased as a function of CSF GFAP and CSF sTREM2.
CSF GAP43 and CSF Ng as a function of CSF GFAP (a) and CSF sTREM2 (b). Lowess curves supported that synaptic dysfunction progressively increased with the increase of astrocyte reactivity biomarker levels across the aging and AD continuum and as a function of microglial activation in the presence of Aβ pathology. Lowess curves were adjusted for age and sex.
Fig. 4
Fig. 4. CSF pTau mediated the association of glial reactivity with synaptic dysfunction.
CSF pTau181 levels fully mediated the association between CSF GFAP and CSF GAP43 in the TRIAD cohort (a–c) and CSF Ng (d–f). In the TRIAD and ADNI cohorts combined, CSF pTau181 levels partially mediated the associations between CSF sTREM2 and CSF GAP43 (g–i) and CSF Ng (j–l), when present. Mediation analysis was conducted using linear regression and the mediate function in R, applying a two-sided test with bootstrapping (500 simulations) to estimate indirect effects. Exact P values are reported where applicable. All models were adjusted for age and sex.
Fig. 5
Fig. 5. Schematic representation of the association between astrocytic and microglial abnormalities with synaptic dysfunction.
Our study revealed that astrocyte reactivity, indicated by increased CSF GFAP levels, is associated with presynaptic (GAP43) and postsynaptic (neurogranin, Ng) markers independently of amyloid-β (Aβ) pathology. Microglial activation, represented by sTREM2, is associated with synaptic dysfunction in an Aβ-dependent manner, showing an association with presynaptic dysfunction (GAP43) only in individuals Aβ+, and with postsynaptic dysfunction (Ng) only when cognitive impairment is present. The schematic highlights key molecular processes involved, including pTau accumulation, Ca2+ signaling, and microtubule disintegration, contributing to neurodegenerative mechanisms. Created in BioRender. Laboratório 28, B. (2025) https://BioRender.com/xek5245.
See this image and copyright information in PMC

Update of

References

    1. Therriault, J. et al. Staging of Alzheimer’s disease: past, present, and future perspectives. Trends Mol. Med.28, 726–741 (2022). - PubMed
    1. Knopman, D. S. et al. Alzheimer disease. Nat. Rev. Dis. Prim.7, 33 (2021). - PMC - PubMed
    1. Gobom, J., Brinkmalm, A., Brinkmalm, G., Blennow, K. & Zetterberg, H. Alzheimer’s Disease biomarker analysis using targeted mass spectrometry. Mol. Cell. Proteom.23, 100721 (2024). - PMC - PubMed
    1. Ansari, M. A., Rao, M. S. & Al-Jarallah, A. Insights into early pathogenesis of sporadic Alzheimer’s disease: role of oxidative stress and loss of synaptic proteins. Front. Neurosci.17, 1273626 (2023). - PMC - PubMed
    1. Vainchtein, I. D. & Molofsky, A. V. Astrocytes and Microglia: In Sickness and in Health. Trends Neurosci.43, 144–154 (2020). - PMC - PubMed

MeSH terms